They trained the AI model to identify waveform patterns in the ECGs that are linked to sudden cardiac death. Most patients flagged by AI as having a risk were missed by the methods doctors currently use.
Video Player is loading.
Current Time 0:19
Duration 1:09
Loaded: 58.11%
Remaining Time 0:50
Researchers at the University of California Berkeley trained an artificial intelligence model to study the results from routine heart tests and look for patterns that indicate a risk for sudden cardiac death.
The model detected signs that doctors have missed for years, according to a study published in Nature.
Sudden cardiac arrest can strike people with health problems, but it also strikes young athletes and people who weren't aware they were at risk, according to Fox News.
The researchers used more than 440,000 electrocardiograms from Sweden and paired the scans with death certificates and health records. They trained the AI model to identify waveform patterns in the ECGs that are linked to sudden cardiac death. They also tested the model on patient data in the U.S. and Taiwan.
Most patients flagged by AI as having a risk were missed by the methods doctors currently use.
Unlock unlimited access
